Journal of ICT, 4, pp: 57-76

A SURVEY ON THE CURRENT PRACTICES OF

SOFTWARE DEVELOPMENT PROCESS IN MALAYSIA

*Fauziah Baharom, ** Aziz Deraman and ** Abdul Razak Hamdan

*Faculty of Information Technology, Universiti Utara Malaysia

060/0 Sintok, Kedah, Malaysia

**Faculty of Technology and Information Science,

University Kebangsaan Malaysia
43600 Bangi, Selangor, Malaysia

Email: fauziah@uum.edu.my
Email: a.d@pkrisc.cc.ukm.my
Email: arh@ftsm.ukm.my

ABSTRACT

Software quality is becoming increasingly important as our society becomes more

dependent on computers. The required levels of quality are not likely to be
achieved without giving proper attention to the software development process.
Many studies regarding the practices of software development process have been
carried out. The focus of these studies is mainly to examine the effectiveness of the
practices. Most of studies were conducted in the United States, European countries,
India, Japan and are very limited in Asian countries especially in Malaysia. In the
view of that, we have carried out a study on the current practices of software
development process in Malaysia. In this paper, we have analyzed and discussed
the results found from the study. The structured questionnaire approach was used
for data collection. The data was analyzed by using basic statistical techniques such
as frequency and cross-tabulation. The results of this study have laid out some
important points regarding the practices of the software development process in
Malaysia. In addition, most of the practitioners were pursuing practices that have
been traditionally used by the organizations.
Lack of awareness in using good software development practices could be a key
factor that contributes to the occurrence of quality

57
 Journal of ICT, 4, pp: 57-76

in the organizations. Consequently, the results of this study have given us a good
problems incentive to prolong our research in the area of software process
certification.

Keywords: Software development practice, software quality, CASE tools and software
testing.

1.0 INTRODUCTION

Nowadays, computers have been recognized as useful tools to support the daily works by almost
all businesses in the world. These tools are commonly used to perform various crucial functions
such as billing, payment, inventory control, purchasing and process control (Ward &
Venkataraman, 1998). As computers are operated by software therefore producing high quality
software is becoming a major challenge for the software industry. Quality was seen as a means to
increase productivity, lower costs and providing the customer with desired product (Wheeler &
Duggins 1998) and lack of quality would give a significant impact on society and economy such
as increasing costs and property damage.

Issues related to software quality problems have been discussed since the early days of
computing. Reports about unfinished development projects, project overruns and system failures
are still the rule (Kautz & Nielsen, 2000). As we know, development of software is known to be a
complex task that covers various activities until the final product can be delivered to the user or
customer. The required levels of quality are unlikely to be achieved without giving proper
attention to the software development process (Bekbay & Liu, 2003). Consequently, several
innovations for improving software development process have been developed and promoted in
the software industry.

Many studies regarding the practices of the software development process have been carried out.
The focus of these studies is mainly to examine on the effectiveness of certain practices. Most of
the studies were conducted in the United States (US), Europe, India and Japan but very few were
carried out in Asian countries especially in Malaysia. Therefore, this study has been carried out to
focus on the current practices of software development process in Malaysia. In addition, the aim
of this study is to investigate the deployment of software development activities, the use of tools,
programming languages and

58
 Journal of ICT, 4, pp: 57-76

techniques, and also to examine the issues and problems that might occur during the development
process. This paper will highlight and discuss the findings found from the study.

2.0 RELATED WORKS

Most of the studies regarding software development practices have been carried out in the United
Kingdom (UK), US, Europe, Japan, India (Wheeler & Duggins, 1998; Cusumano et al, 2003) and
New Zealand (Groves et al, 2000). However, very limited numbers of similar research studies
have been carried out in Asia and particularly in Malaysia (Yazrina et al, 2002).
Ward and Venkataraman (1998) had observed on various approaches for improving software
quality such as better evaluation techniques, better quality measurement, better development
processes and better development tools.

Blackburn, Scudder and Van Wassenhove (1996) conducted a survey, which aimed to identify the
factors that contributed to the timeliness of software development. Several factors were identified
which are the use of prototyping approach, CASE tools, parallel development, communication
between team members, reuse of code and implementation of testing strategies. The results also
indicated that spending more time and effort on customer specifications will improve the
development productivity and speed.

In 1990, Cusumano and Kemerer (1990) published a survey of 40 software development projects
in the US and Japan. The results showed that both countries were similar in several areas such as
types and size of software products developed, development tools utilized, programming
languages and hardware platforms used for development. However, they were differ in term of
time spent during software development projects. In Japan, the study showed that their software
people spent more time on product designing phase while in the US, more time was spent on
actual coding. As a result, Japanese projects were completed with fewer numbers of defects and
higher line-of-code productivity compared to the American projects.

Starting from 2001, the study on software development practices was prolonged to be more
global. Cusumano, MacCromack, Kemerer and Crandall (2003) had conducted a global survey of
software development practices. Their study

59
 Journal of ICT, 4, pp: 57-76

took about 2 years to complete. A total number of 104 software development projects responded
to the survey. The projects were primarily centered in India, Japan, Europe and the US. Most of
the software projects in India, Japan and Europe used architectural, functional and design
specifications rather than just writing code with minimal planning and documentation. This is
different compared to software projects in the US, where the sample indicated that specifications
are less frequently used. The sample also reported that 100% of Indian projects used detailed
designs and implemented design reviews. These responses suggest that there is some truth to the
recent speculation that Indian software companies are increasingly adopting more formal project
management techniques. Almost all of the samples made use of early beta releases, which is
becoming a useful tool for testing and obtaining user feedback. In Japan, the waterfall process is
still a popular choice, while most of the Indian, European and US projects are broken into
multiple milestones known as the incremental process.

Yazrina etal (2002) have done a study on the use of system development methodology in
Malaysia. The method used to carry out the study is by using questionnaire. The results of the
survey have been analyzed using basic statistical measures such as mean comparison, frequency,
cross tabulations, correlation and variance analyses. Findings showed that there were some
respondents who did not use any methodology while developing an information system due to
lack of expertise and too many deliverables that need to be delivered from the methodology.
However, 43.3% of the respondents agreed that the use of methodologies are helpful in the
development of an information system because it involved certain type of tools such as CASE
tools, word processing, spreadsheets, graphic and presentation software. Tools ease the
information system development and the development could be completed on schedule or time.

Groves et al (2000) had conducted "A Survey Of Software Development Practices in the New
Zealand Software Industry". The study focused on determining the techniques and tools used for
expressing software requirements. The group of researchers reported that there is a wide range of
kinds and sizes of software organizations as well as type of applications developed in New
Zealand. This diversity makes it difficult to identify meaningful patterns in the software
development practices employed. The larger software development groups tend to have more
well-defined software development processes, spending more time on capturing requirements and
to have more rigorous testing regimes. It

60
 Journal of ICT, 4, pp: 57-76

also appears that the well-defined software development processes are used to produce software
for customers rather than in-house providers. Most of the respondents used text documents and
general-purpose tools such as word processors and spreadsheets to express requirements and
specifications. Most companies stated that they used standard change request forms to capture
requirements changes and they hold review meetings after each early lifecycle phase except for
code reviews.

3.0 RESEARCH METHOD

The general objective of this survey is to examine the current practices of software development
processes and to identify the related issues and problems which commonly occur in producing
software products in Malaysian organizations. It is supported by the following six specific
objectives:
i. To determine the extent of the use of standards and procedures, tools, methods and software
development life cycle models
ii. To determine the deployment of software development activities
iii. To determine the extent of software reuse
iv. To determine the awareness of the user or/and customer involvement during software
development
v. To identify the quality problems whichcommonly occur in producing software product
vi. To determine the awareness of improving the developer's skills

A structured questionnaire with 69 questions was used in this study. The questionnaire was
divided into four sections, which include:
i. Respondent background. This section covers the general information of the respondent
such as the current position, years of experience, his/her involvement in information
technology and activities in which he/she is currently involved.
ii. Organizational profile. It solicits information regarding the type of
organization, organization's function or type of industry, whether the
organization has an IT department and number of IT professionals in the
organization.
iii. Process quality control. This section seeks information related to the use of life cycle
models, the deployment of standards and procedures,

61
 Journal of ICT, 4, pp: 57-76

the use of methodologies, tools and techniques and the implemented activities during
development.
iv. Human resource development. This last section looks for trainin g programs
provided by the organization for improving the skill of the staff.

4.0 DATA ANALYSIS

About 220 sets of questionnaires were distributed to private and public sectors. Out of this
amount, only 45 were returned, 4 sets without answers and only 41 sets were answered
completely and were used for analysis. The results of the survey have been analyzed by using
basic statistical methods, which are frequency and cross tabulations. These will be discussed in
the next section.

4.1 Demographics and general information

The set of questions asked in the first two sections of the questionnaire are used to assess the
qualifications of the respondents and the organization profile.
 Respondents' Qualifications
In this section, the respondents were asked about their current positions and years of
experience. The data was then analyzed using cross tabulations analysis. Table 1 shows the
results obtained from the analysis.

Out of the 41 respondents, only 26.83% had working experiences of more than 10 years and most
of them were Systems Analysts.

62
 Journal of ICT, 4, pp: 57-76

• Organizational Profile

The second section of the questionnaire is needed in order to study the organizational background.
Table 2 depicts the number of organizations that have an IT unit or department.

From the results acquired, it can be concluded that most organizations have IT departments
especially the government agencies (100%). On the other hand, only 4 (20%) out of 20 private
agencies do not have IT departments. These organizations are involved in different types of
businesses such as Education and Training, Construction, Software development and so on. Table
3 shows the distribution of the organizations based on their type of business.

63
 Journal of ICT, 4, pp: 57-76

A high percentage of respondents came from small IT-based organizations and the number of IT
professionals employed is between three to five people (see Table 4).

4.2 Software Development Practices

In this section, the results of the findings are presented based on the stated objectives.

Objective 1: To determine the extent ofthe use of standards and procedures, tools, methods
and software development life cycle models.

• The use of standards and procedures

The use of standards and procedures is required to bring uniformity and control to the process of
developing a software product (Wheeler & Duggins, 1998). In this survey, our respondents were
asked about the standard to which they refer such as IEEE, ISO 9000, CMM, other international,
local standards or none at all.

From the analysis, the results indicated that 36.59% of the organizations that responded are
following their own standards. Surprisingly, 31.71% of organizations are still unaware of the
importance of standards in producing high quality products. Only 7.32% are using IEEE standards
and none of them are using CMM (0%) to control their software development process. ISO 9000

64
 Journal of ICT, 4, pp: 57-76

standards are the most acceptable international standards in Malaysia. The detail results obtained
are as shown in Table 5.

The results also convey that formal procedures are used to:
 manage the software development project such as cost estimation and project scheduling
(80.5%).
 ensure a deliverable is properly understood whenever it is passed from one discrete group
to another (e.g. user to analyst to designer) (80.5%).
 control changes to software requirements, designs and other documentation during
development (75.6%).
 check whether common code standards were applied to each software project (63.4%).
 ensure that formal reviews of project status were periodically conducted (63.4%).
 make sure that independent audits (such as inspections or walkthroughs) were conducted
for each major stage in the software development process (48.8%).

 The Use of CASE Tools

Nowadays, many CASE tools are available that can be used to support the software development
process such as management tools, editing tools, prototyping, supporting tools, testing and
debugging tools. These tools are helpful as they lead to some improvement in software quality
and productivity.

From the analysis, it is shown that private sectors are more aware in using CASE tools compared
to government agencies. The results are presented in Figure 1.

65
 Journal of ICT, 4, pp: 57-76

The use of specific notations is helpful in presenting software requirement specifications. In

addition, it will create better communication among developers or between developers and
users/customers. From the results acquired, most of the organizations preferred to use semi-formal
notations (e.g. DFD, ERD) compared to formal (e.g. Z, VDM) and informal (Natural Language)
notations.

Table 6lists the various programming languages used by the respondents during development
activities. It is obvious from Table 6 that visual language and object oriented are the most popular
languages used by the organizations.

• The Use of Software Process Model

Normally, specific software process models such as Waterfall, spiral, rapid prototyping will be
used to develop software. In this study, the respondents

66
 Journal of ICT, 4, pp: 57-76

were also required to specify the level of software process model that they used. From the results
attained, 39.02% of the organizations had never used spiral and object-oriented models. The most
acceptable model that was regularly and always used is the waterfall model. Table 7 below
presents the detailed results:

• The Use of Requirement Collection Techniques

The most critical activity that always arises during software development is the collecting
requirement of the proposed system. This activity is vital to have a clear understanding on the
needs of user or/customer. Understanding the nature of the problems can be very difficult
especially if the software system is new. Therefore, in order to simplify the requirement collection
process, some techniques and methods have been introduced.

Table 8 specifies the various techniques that can be used for requirement collection. In this study,
interview seems to be the most popular technique that has always been used by the organizations.

67
 Journal of ICT, 4, pp: 57-76

The various software requirement analysis methods used by the organizations have been listed in
Table 9. The table also reveals that the Jackson System Development method has not been applied
by any of the organizations. More than half of the organizations are using SAD, SSADM and
OOA methods.

• The Deployment of Software Development Activities

This section will discuss the deployment of activities during software development. From the
results acquired, half the organizations commonly used informal and ad-hoc methods and
procedures with reasonable control of schedules and cost. Software measurement is an activity
used to understand, improve and predict the quality of the software process or product. The survey
indicates that not more than 24.4% of the organizations have implementing quantitative
measurement in order to improve the quality of their product.

Total effort spent can be used to measure the effectiveness of a software development process. By
looking at Table 10, it is hard to find an organization that spends more than 20% of their efforts
for planning and requirement analysis. Most of the organizations spent from 5% to 20% of their
efforts for planning and requirement analysis.

68
 Journal of ICT, 4, pp: 57-76

Testing is an important process in improving the quality of the software product. The purpose of
this process is to find errors, which might occur during specification, design and code generation.
Based on the results acquired, formal testing is widely implemented in government and private
organizations. The results obtained are as follows:

The results depicted in Table 12 indicate that most of the organizations began to apply formal
software testing at the end of the coding phase (41.5%). The rest of the results are as follows:

Table 13 indicates several available testing techniques commonly practiced during software
development. The results show most of the organizations carried out unit testing, integration
testing, system testing and acceptance testing.

69
 Journal of ICT, 4, pp: 57-76

However, about half of the organizations were practicing alpha and beta testing.

The results also indicate that while carrying out a software project, 58.5% of the organizations
were producing formal test plans. In additional, 73.2% of the organizations agreed that software
testing by itself is not sufficient for guaranteeing the software product meets with its
requirements.

Objective 3: To examine the extent of software reuse

Software should be considered as an asset and reuse of these assets are essential to increase the
return on their development costs. Demands for lower software production costs, faster delivery
and increased quality can only be met through widespread and systematic software reuse
(Sommerville, 200 1). From the results obtained, most of the organizations are providing good
encouragement on the creation and implementation of software reuse.

For the organizations that encourage reuse, the most three ranked components that are frequently
being reused were source code (65.85 %), design documents (48.78%) and user
documentation/specification (48.78%).

70
 Journal of ICT, 4, pp: 57-76

In the degree of code modification, the results also indicate that only 9.76% of the organizations
reuse the source code totally without any changes and 46.34% of the organizations had changed
more than 20% of lines code. The detailed results are indicated in Table 16 below:

In the trend of improving software quality, involvement from the user or/and customer is
considered as essential. From the results revealed in Table 17, both the government and private
sectors are concerned about this issue. More than half of the organizations stated that the user
or/and customer is regularly involved during the software development process.

71
 Journal of ICT, 4, pp: 57-76

We continued our investigation by looking for the quality problems that commonly occurred in
the organizations. In our questionnaires, we had listed four types of quality problems which are:
software can not be delivered on time, over budget, customer unhappy with the product quality
and software needs further improvement or maintenance after it was released. From the analysis,
all of the organizations claimed that they have faced with at least one of the quality problems and
the results are classified as follows:
 75% of the organizations mentioned that their software products need further improvement
after they were released
 55% of the organizations could not deliver their software systems on time
 20% of the organizations faced with budgetary problems
 22.5% of the organizations were confronted by their customers who were unhappy with the
quality of the software products.

Objective 6: To find out the awareness of improving the developer's skills.

Training of employees is an important item. Almost all of the organizations (80.49%) claimed that
they provide their developers and project managers training in improving software development
or project management skills. More than half of the government agencies and 45% of the private
sectors were providing training from I to 2 times per year. The table below shows the detailed
results about the number of training seasons provided annually by the organizations.

72
 Journal of ICT, 4, pp: 57-76

The results also indicate that 97.6% of the organizations agreed that know ledge and skills could
be improved through attending appropriate training, courses and/or seminars. Training, experience
and formal education are three factors that have been identified which can contribute in improving
knowledge and skills in software development. From the analysis, we found majority of the
organizations believed that experience is the most important factor. The result is depicted in
Figure. 2.

5.0 CONCLUSION

It was discovered that most of the respondents work as system analysts and a major percentage of
them had less than 3 years of experience. Out of 41 organizations, only 11 organizations were
totally involved in software development. Looking at the respondents' years of experience (Table
1) and number of IT professionals (Table 4), to produce high quality software, standards,

73
 Journal of ICT, 4, pp: 57-76

CASE tools and methodologies should be used extensively. However, Table 5 indicated that
31.71 % of the organizations are still not referring to any standards when developing software. A
number of CASE tools were introduced to support the software development process therefore the
expected software product could be delivered on time. Unfortunately, Figure 1 depicts that in the
government agencies the use of CASE tools appeared to be very small especially for prototyping
(23.81 %), testing (19.05%) and debugging (19.05%).

According to Pressman (2001) more than 40 percent of all efforts is often recommended for
analysis and design for large software development projects. However, from the results obtained
in Table 10, only a small percentage of the respondents were following this recommendation.
Producing high quality software, economically and with fast delivery are supposed to be the goals
of every software developer. To achieve these goals, use of methods during development process
is certainly important. However, the results in Table 9 shows that software requirement analysis
methods are not been applied by every organization due to reasons such as lack of expertise and
too many deliverables (Yazrina et al, 2002). Therefore, for a method to be utilized, the
organizations should allocate sufficient training for their staffs. In addition, adopting object and
reuse oriented process models are also necessary. From the results obtained in Table 7, most of
these process models have never been used. The private sector however was more encouraging in
software reuse development (Table 14) due to intention to compete with their competitors.

Consequently, the results of the study have laid out some important points to be considered such
as:
 It was discovered that most of organizations in Malaysia were still lacking in the use of
standards.
 The majority of the organizations were facing problems such as quality, late delivery, over
budget and software products needing further improvement or modification.
 The use of CASE tools in the government agencies (67.7%) appeared to be not so
encouraging compared to the private sector (85%).
 34.15% of the organizations did not use any specific method for performing requirement
analysis activities.
 Source code was the only component that was frequently being reused by the organizations.
Therefore, the reuse of other components should be considered in improving software
quality.

74
 Journal of ICT, 4, pp: 57-76

 For the deployment of process model, the waterfall method was more acceptable compared
to other models such as object oriented and reuse oriented.
 There were a small number of organizations that spent more than 20% of their effort for
executing planning and analysis activities.
 Most of the organizations were initially executing formal software testing after coding
phase were finished. Actually it is also possible to do this formal testing earlier, for
instance, as soon as software projects were creating documentations or elements that could
be tested.
 Alpha and beta testing was hardly implemented in the organizations.

To conclude, most of the practitioners were pursuing practices that have been traditionally used
by the organizations. Lack of awareness in using good software development practices could be a
key factor that contributed to the occurrence of quality problems in the organizations. The results
of this study have given us a good incentive to prolong our research in the area of software
process certification.

6.0 ACKNOWLEDGEMENT

This research is funded by the Intensification of Research in Priority Areas (IRPA) from Malaysia
Ministry of Science, Technology and Environment (MOSTE).

7.0 REFERENCES

Bekbay, S. & Liu, S. (2003). A Study of Japanese Software Process Practices and a Potential for
Improvement Using SOFL. IEEE Computer Society, 100 - 107.

Blackburn, D. J., Scudder, D. G. & van Wassenhove, N. L. (1996). Improving Speed and
Productivity of Software Development: A Global Survey of Software Developers. IEEE
Transactions on Software Engineering, 22(12), 875-885.

Cusumano, M. & Kemerer, F. C. (2003). A Quantitative Analysis of US and Japanese Practice

and Performance in Software Development. Management Science, 36(11), 1384-1406.

75
 Journal of ICT, 4, pp: 57-76

Cusumano, M., MacCormack, A., Kemerer, F. C. & Crandall, W. (2003). A Global Survey of
Software Development Practices. IEEE Software.

Groves, L., Nickson, R., Reeve, G., Reeves, S. & Utting, M. (2000). A Survey of Software
Development Practices in the New Zealand Software Industry. IEEE Software.

Kaurtz, K. & Nielsen, A. P. (2000). Implementing Software Process Improvement: Two Cases of
Technology Transfer. Proceedings of the 33rd Hawaii International Conference on System
Sciences 2000.

Pressman, R. S. (2001). Software Engineering: A practitioner's Approach. McGraw-Hill.

Sommerville, 1. (2001). Software Engineering (6th ed.): Addison Wesley.

Wheeler, S. & Duggins, S. (1998). Improving Software Quality. ACM, 300309.

Ward, A. W. & Venkataraman, B. (1999). Some observations on software quality. A CM

Southeast Regional Conference 1999.

Yazrina, Y. (2002). The Use ofInformation System Development Methodology in Malaysia.

Jurnal Antarabangsa (Teknologi Maklumat), 2(2002), 15-34.

76